Mark Salyzyn writes:
>Christoph Hellwig writes:
>>> + aac_driver_template.name = aac_drivers[index].name;
>>this is broken. it overwrites the name every time a host is allocated.
>Especially it does this on a live template. Just implemented ->info
>
>->info() is not enough, there are places in the scsi subsystem which do
>not implement the info override over top the template name.
>
>I had discussed the possibility of submitting a patch to the scsi
subsystem to fix that brokenness, I am generally concerned with legacy
issues.
Looks like few remain. Could not figure a way to alter scsi_module.c
that made sense, doable but expensive, but hosts.c could be fixed up
with:
--- hosts.c 2005-09-07 12:17:18.000000000 -0400
+++ /tmp/hosts.c 2005-09-12 09:57:52.906881694 -0400
@@ -166,7 +166,7 @@
if (!shost->can_queue) {
printk(KERN_ERR "%s: can_queue = 0 no longer
supported\n",
- sht->name);
+ sht->info ? sht->info(shost) :
sht->name);
goto out;
}
@@ -386,7 +386,8 @@
if (!sht->detect) {
printk(KERN_WARNING "scsi_register() called on new-style
"
- "template for driver %s\n",
sht->name);
+ "template for driver %s\n",
+ sht->info ? sht->info(shost) :
sht->name);
dump_stack();
}
@@ -483,7 +484,8 @@
if (unlikely(!shost->work_q)) {
printk(KERN_ERR
"ERROR: Scsi host '%s' attempted to queue
scsi-work, "
- "when no workqueue created.\n",
shost->hostt->name);
+ "when no workqueue created.\n",
+ shost->hostt->info ? shost->hostt->info(shost) :
shost->
hostt->name);
dump_stack();
return -EINVAL;
@@ -502,7 +504,8 @@
if (!shost->work_q) {
printk(KERN_ERR
"ERROR: Scsi host '%s' attempted to flush
scsi-work, "
- "when no workqueue created.\n",
shost->hostt->name);
+ "when no workqueue created.\n",
+ shost->hostt->info ? shost->hostt->info(shost) :
shost->
hostt->name);
dump_stack();
return;
Sincerely -- Mark Salyzyn
-
: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
